说明：  最优的阶梯型边缘检测算法(canny边缘检测) 1.Canny边缘检测基本原理 (1)图象边缘检测必须满足两个条件：一能有效地抑制噪声；二必须尽量精确确定边缘的位置。 (2)根据对信噪比与定位乘积进行测度，得到最优化逼近算子。这就是Canny边缘检测算子。 (3)类似与Marr（LoG）边缘检测方法，也属于先平滑后求导数的方法。 2.Canny边缘检测算法： step1:用高斯滤波器平滑图象； step2:用一阶偏导的有限差分来计算梯度的幅值和方向； step3:对梯度幅值进行非极大值抑制； step4:用双阈值算法检测和连接边缘。 step1:高斯平滑函数
(The optimal type stair edge detection algorithm (canny edge detection) 1 Canny edge detection principle (1) the image edge detection must meet two conditions: one can effectively reduce the noise, 2 must try to accurately determine edge position. (2) according to the signal-to-noise ratio and the product localization estimate, get optimal approximation operator. This is the Canny edge detection operators. (3) similar to Marr (LoG) edge detection method, also belongs to the first derivative method for smooth after. 2 Canny edge detection algorithm: Step1: using gauss filter smooth image, Step2: a partial derivatives with finite difference to calculate gradient value and direction, Step3: a gradient of maximum inhibition, Step4: double threshold algorithm of edge detection and connection. Step1: gaussian smooth functions)